Skip to content

asf_search Basics#


asf_search is a Python package for performing searches of the ASF catalog. In addition, it offers baseline functionality and download support. It is available through PyPi and Conda.

import asf_search as asf

results = asf.granule_search(['ALPSRS279162400', 'ALPSRS279162200'])

wkt = 'POLYGON((-135.7 58.2,-136.6 58.1,-135.8 56.9,-134.6 56.1,-134.9 58.0,-135.7 58.2))'
results = asf.geo_search(platform=[asf.PLATFORM.SENTINEL1], intersectsWith=wkt, maxResults=10)

For an introductory walkthrough of asf_search, see the Jupyter Notebooks.


In order to easily manage dependencies, we recommend using dedicated project environments via Anaconda/Miniconda or Python virtual environments.

asf_search can be installed into a conda environment with

conda install -c conda-forge asf_search

or into a virtual environment with

python -m pip install asf_search


Programmatically searching for ASF data is made simple with asf_search. Several search functions are provided. Each search function returns an ASFSearchResults object:

  • geo_search() Find product info over an area of interest using a WKT string
  • granule_search() Find product info using a list of scene names
  • product_search() Find product info using a list of product IDs
  • stack_from_id() Find a baseline stack of products using a reference scene
  • If the above search approaches do not meet your search needs, search() supports all available keywords:
    • search() Find product info using any combination of search parameters
  • Additionally, numerous constants are provided to ease the search process. Currently, we provide constants for beam mode, flight direction, instrument, platform, polarization, and product type. You can see the full list of constants here.

Additionally, asf_search supports downloading data, both from search results as provided by the above search functions, and directly on product URLs. An authenticated session is generally required. More information on available authentication methods can be found here. You may also authenticate using an ASFSession object and one of the following authentication methods. ASFSession is a subclass of Session.

  • auth_with_creds('user', 'pass)
  • auth_with_token('EDL token')
  • auth_with_cookiejar(http.cookiejar)

If not using .netrc credentials, that session should be passed to whichever download method is being called, can be re-used, and is thread safe.

Example using .netrc:

results = ....'....')

Example with manual authentication:

results = asf_search.granule_search([...])
session = asf_search.ASFSession().auth_with_creds('user', 'pass')'/Users/SARGuru/data', session=session)

Alternately, asf_search supports downloading an arbitrary list of URLs. All of the available authentication methods are supported:

urls = [...]
asf_search.download_urls(urls=urls, path='/Users/SARGuru/data', session=ASFSession().auth_with_token('EDL token'))

Also note that and the generic download_urls() function both accept a processes parameter which allows for parallel downloads.

Further examples of all of the above can be found in this sample script.